Warwick-Plasma / epoch

Particle-in-cell code for plasma physics simulations
https://epochpic.github.io
GNU General Public License v3.0
186 stars 59 forks source link

no SDF files output #561

Closed TrumeAAA closed 1 year ago

TrumeAAA commented 1 year ago

Hi, I find that there are no sdf files generated. The err file reads: whoami: cannot find name for user ID 1030 srun: fatal: Invalid user id: 1030 However, when I run the simulation with root, it works again. What happened? Any suggestions will be helpful. Ai,

Status-Mirror commented 1 year ago

Are you trying to write to output SDF files to a directory you don't have permission to write to? This would explain why root privileges would work.

Cheers, Stuart

TrumeAAA commented 1 year ago

Hi, Thank you for your reply. It might be a reason that I don't have permission to write. So I tried to change permission by run " chmod 777 laser ". And I get the permission as can be seen: 屏幕截图 2023-10-06 192501 However, the same error still exist: 屏幕截图 2023-10-06 192549 Ai.

Status-Mirror commented 1 year ago

I haven't seen this error before, and I'm not sure what would be causing it.

I'm including a small input deck here which should produce two SDF files in under a second on one core - this should be small enough to run without queueing on your cluster. If you find you have a problem running this one, then I would contact your local admin team (this seems like a cluster problem instead of an EPOCH one).

begin:control
    nx = 200
    ny = 100
    t_end = 10e-15
    x_min = 0
    x_max = 10e-6
    y_min = -2.5e-6
    y_max = 2.5e-6
    stdout_frequency = 100
end:control

begin:boundaries
    bc_x_min = simple_laser
    bc_x_max = open
    bc_y_min = open
    bc_y_max = open
end:boundaries

begin:laser
    boundary = x_min
    intensity_w_cm2 = 1.0e15
    lambda = 1.0e-6
    profile = 1
end:laser

begin:output
    name = o1
    dt_snapshot = 10 * femto
    poynt_flux = always
end:output

Hope this helps, Stuart

TrumeAAA commented 1 year ago

Hi, Sorry for reply so late. I think it is a cluster problem because Epoch can work normally after restarting the cluster though I haven't figure out what happened( sorry for that ). Thanks for you kind reply! cheers, Ai